Jakarta, INTI – OpenAI has once again introduced a breakthrough innovation by launching a more advanced version of its artificial intelligence (AI) reasoning model, o1. The latest model, named o1-pro, is claimed to generate more accurate and consistent responses thanks to increased computational power.
Availability and Pricing
According to a TechCrunch report on Wednesday, March 19, 2025, the o1-pro model is currently available only to select developers who have spent at least $5 (approximately IDR 82,600) on OpenAI's API services. However, the cost of using this model is relatively high. OpenAI charges $150 (approximately IDR 2.4 million) per one million input tokens (about 750,000 words) and $600 per one million output tokens generated by the model. This cost is twice as high as GPT-4.5 for input and ten times more expensive than the standard o1 model.
Performance Improvements and OpenAI’s Expectations
This price increase reflects OpenAI’s confidence in the superior performance of o1-pro. A spokesperson for OpenAI stated that the model is designed to "think harder" and provide better answers to complex problems.
"After receiving numerous requests from the developer community, we are excited to bring this model into the API to offer more reliable responses," said the OpenAI spokesperson.
Challenges and Criticism of o1-pro
However, despite promising performance improvements, initial responses to o1-pro have not been entirely positive. The model, which has been available on OpenAI's ChatGPT platform for ChatGPT Pro subscribers since December 2024, has reportedly struggled with solving Sudoku puzzles and has been deceived by simple optical illusion-based jokes. Additionally, OpenAI’s internal benchmarks at the end of last year indicated that o1-pro’s performance improvement over the standard o1 model in solving programming and mathematical problems was not very significant, although its answers were more reliable.
Conclusion
With various advantages and challenges still present, the o1-pro model remains an important step forward in AI reasoning development. OpenAI hopes that these performance enhancements will be sufficient to attract developers who require more advanced and reliable AI solutions.
Read More : Baidu Challenges DeepSeek: Advanced AI at a Lower Cost